I build software and lead
the teams behind it.
I'm Kads Aziz: engineer, then lead, then founder, now advisor. I write about the parts of software delivery people skip: judgement, accountability, leadership, and regulation, and what happens to all four when a team adopts AI. Most of my work lives in the gap between what a business wants to build and what a system can safely support.
Background
My career runs across e-commerce, healthcare, and aged care, most of it regulated. I've built products, led engineering teams, and shipped the software below. I've also untangled stuck delivery, modernised legacy systems, and chaired the awkward conversations where product, compliance, engineering, and operations collide.
All of it comes from doing the job, not watching it.
What I've shipped
The clearest proof isn't the writing. It's the software with my name on it, live and in other people's hands.
EngLedger
AI spend governance and a finance-grade engineering ledger. It meters every AI request across providers and ties it to the work it produced, then reconciles engineering effort into numbers an auditor can trace back to evidence. It's the argument behind most of my writing, built into software: AI that makes a team faster without making its cost or its decisions impossible to explain.
Nimbo
An Australian ecommerce platform for small businesses: storefront, payments, orders, and hosting in one place. It has been live for years, running real shops and real money, which teaches you the side of software a demo never shows: uptime at 2am, a support queue with a real person on the other end, and a checkout that can't go down mid-sale.
What I write about
AI adoption with accountability
The throughline of most of what I write. AI makes software teams faster; it doesn't make them less responsible. In regulated work it usually makes accountability harder to place, not easier. I write about where AI earns its place, where it doesn't belong, and what leaders need to adopt it well: human judgement that stays in the loop, audit trails that hold up, and systems you can still explain. That extends to evidence that survives a reviewer's question, whether the reviewer is an auditor or the ATO. Most of it lives in the Still Accountable series.
Engineering leadership
Leading software teams without the performance of managing them. Clarity over control, holding a standard without micromanaging, honest delivery signals instead of green dashboards, and an operating rhythm that lets good people do their best work without guessing at priorities.
Regulated software delivery
Regulated software isn't normal software with extra paperwork. The stakes are higher and the failure modes are worse, so the habits have to be different. I write about the delivery choices that keep trust intact in health, aged care, and government-adjacent systems: governance, compliance, and controls that survive inspection.
Building products
Years of building and running products. I write about the unglamorous parts that decide whether a product lasts: positioning, technical debt, the promises you made to customers, and the small decisions nobody celebrates.
Work with me
Open to a few engagements
I take on a small number of advisory and fractional engagements with Australian teams, mostly in health, aged care, government-adjacent software, SaaS, and other regulated environments. The work that fits me best is the kind that's technical, organisational, and commercial at once.
The questions usually sound like:
- Where should AI actually go in our delivery, and where shouldn't it?
- How do we adopt AI without losing track of who's accountable?
- How do we see delivery risk earlier, before it gets expensive?
- Who leads engineering while we find the right person?
- How do we lead engineering better without piling on process?
- How do we make changes to a legacy system safer and more predictable?
- How do we keep product intent, engineering, and auditability pointing the same way?
If what you need is something built rather than advised on, that's Buildlight Labs, the studio I co-founded.
A two-week delivery review
I spend two weeks inside your engineering: talking to the team, reading the code and the pipeline, looking at how work gets planned, shipped, and audited. You get a written report on where the risk sits, where AI belongs and where it doesn't, and what to fix first, in order. Fixed scope, fixed fee, and the report is yours to act on with or without me.
If that's what's on your desk, [email protected], or leave it here.